This RO1 resubmission is an application in response to PA-04-100 encouraging research on the origins and pathways to substance abuse (SA). The adult literature indicates that SA is strongly associated with bipolar disorder (BPD). An emerging body of evidence suggests that juvenile BPD may place adolescents at high risk for SA and that adolescents with SA may be at higher risk to have BPD. We have completed baseline assessments in a NIDA funded family study investigating the relationship of SA in juvenile BPD. Using structured psychiatric interviews, objective and subjective measures of SA, a neuropsychological battery, and measures of psychosocial functioning, we have completed comprehensive evaluations on 110 adolescents with bipolar disorder, their siblings and parents;and a control group of 110 non-mood disordered adolescents, their siblings and parents. We now propose a four-year follow-up of this cohort as they pass through the age of risk for SA. Our baseline data is very encouraging in disentangling the risk of SA in BPD. Data analyses from 76% of the families collected reveals that adolescents (mean age 13.5 yrs) with BPD are at higher risk for SA than adolescents without BPD (30% vs 4%, p<0.001) that is not accounted for by conduct disorder. Adolescent-onset BPD (age >12), relative to child-onset BPD, is a greater risk for SA. BPD youth have significantly greater psychiatric comorbidity, neuropsychological impairment, and psychosocial distress, all independent risk factors for later SA. We propose a four-year follow-up of our sample of control and BPD families to evaluate the risk for cigarette, SA, stressors, and associated dysfunction. We will test hypotheses corresponding to four main aims: 1) Evaluating BPD as a risk factor for SA, 2) Characterizing the course and outcome of SA in BPD youth, 3) Evaluating predictors, mediators, and moderators of SA in BPD adolescents and young adults, and 4) Evaluating the course and outcome of high risk siblings of BPD probands. The proposed work is a critical step for several reasons. This is the first, prospective, family-based study of SA in juvenile BPD;no studies have disentangled risk factors in juvenile BPD leading to SA. Since BPD is identifiable and treatable, findings from the four-year follow-up will provide valuable information to formulate prevention and early intervention strategies.